Came across Walkscore yesterday. It is a Google Maps mashup which takes an address, plots out all of the restaurants, grocery stores, and other common personal services within "walking" distance, and then scores the location on a scale of 0 through 100 for walkability. Pretty neat, not surprisingly the places I've lived or am living in in Brooklyn and Manhattan score well into the "90s". There's some glitches though, which I'll describe after this table I whipped up of the walkability of places I've lived in the past 40 years (ignore the fact that as a one year old I was unlikely to walk anywhere):

AddressYearsWalkscore
North Ave, Oak Park, IL1967-196963
Wilcox St, Downers Grove, IL1969-197151
Brainard St, Naperville, IL1971-197578
Green Valley Circle, Culver City, CA197580
Burr Oak Ln, Lisle, IL1975-197658
Carpenter St, Downers Grove, IL1976-198538
Randolph St, Meadville, PA1985-198982
Wightman St, Pittsburgh, PA1989-199174
Greenhill Dr, Fishkill, NY1991-199223
Theresa Blvd, Wappingers Falls, NY1992-19940!
Hudson Harbor Dr, Poughkeepsie, NY1994-199555
Smokerise Ln, Wappingers Falls, NY1995-1997NA
Battery Place, New York, NY1997-200097
Bridge St, Sydney, NSW, Australia2000NA
Hicks St, Brooklyn Heights, NY, NY2001-200795
Main St, Brooklyn, NY2007-95

Glitches

I assume all of these glitches have more to do with the Google Maps API than the "patent-pending" Walkscore algorithm:

  • I tried various iterations of the address on Smokerise Lane but Walkscore (really the Google Maps API) couldn't geocode it. It wasn't exactly new when I lived there ten years ago, I don't know enough about the API to guess where it's breaking.
  • It totally blew my Sydney address on two levels: it placed it in North Sydney (which is a separate city, I think), and because Google do not have local data for Sydney (again: guessing) it rated it a 0 out of 100. If there's no applicable data at all something (Google Maps API? Walkscore?) should flag that as the condition and not just hand out a 0. Not that anyone who'd solely judge where to live based on Walkscore and thus rule out the Sydney CBD deserves to live in Sydney in the first place. I'm guessing that Google only has local data for the US.
  • For our current place on Main Street in DUMBO, it includes services across the river in Manhattan. Again, knowing nothing about the Google Maps API but feeling the need to opine nonetheless, something, somewhere should be able to clue in and say Yes, these things are within a mile radius of the starting point, however there is a (bridge|river|landfill|mountain) in between. This is not a new problem, plotting out the closest Staples Store to Amagansett, NY returns two locations in Connecticut which are technically closer than the third choice in Riverhead, NY, but require either swimming or flying to reach.

Still, a very cool mashup demonstrating the power of open APIs and open data for that matter.

Via: echo.

«Sailor | Main | Test your AQ»

Comments

EdGy added:

Hey Ed,

Another glitch that I doubt they have the data to do anything about is taking account of lack of sidewalks.

When I lived in Danbury, CT, sure it was possible that I could walk many of the places it highlighted. The idea of walking on the shoulder of Federal Rd, a 4 lane "road" with a speed limit of 45 is something to avoid.

Also noticed that a large residential neighborhood lowers your score as well. Where I grew up in Queens only scored a 71. I think the walkability score would prefer you to live next to or on top of a strip mall or something. :)

83rd Pl, Middle Village, NY 71
Old Brookfield Rd, Danbury, CT 69
Langdon St, Cambridge, MA 92

- EdGy

Slightly acerbic and eccentric dog walker who masquerades as a web developer and occasional CTO.

Spent five years running the technology side of the circus known as www.ibm.com.

More about me here.

Archives

Get updates via email

Enter your email address:

Delivered by FeedBurner